Setup & Installation
What This Skill Does
Deploys container-based agents to Azure AI Foundry using `ImageBasedHostedAgentDefinition` from the Azure AI Projects SDK. You provide a container image and resource specs; the SDK handles registration, versioning, and protocol configuration. Supports tools like code interpreter and MCP servers.
Instead of manually configuring Azure container hosting, capability hosts, and protocol wiring through the portal, this SDK call handles agent registration and versioning in a few lines of Python.
When to use it
- Deploying a custom Python agent as a versioned container in Azure AI Foundry
- Registering a data-processing agent image with specific CPU and memory limits
- Managing multiple versions of a hosted agent and rolling back to a previous one
- Attaching code interpreter or MCP tools to a container-based agent definition
- Passing secrets and model endpoint config to a running agent container via environment variables